the goms family of user interface analysis techniques : comparison and contrast
DESCRIPTION
The GOMS Family of User Interface Analysis Techniques : Comparison and Contrast. Bonnie E. John David E. Kieras Young-joo Jeon. Abstract Introduction Comparison (1.2.3.4.) Summary and Comparison (1.2.3.) Conclusion. C ontents. A bstact. - PowerPoint PPT PresentationTRANSCRIPT
마스터 제목 스타일 편집
마스터 텍스트 스타일을 편집합니다둘째 수준셋째 수준넷째 수준
다섯째 수준The GOMS Family of User Interface Analysis Techniques
: Comparison and ContrastBonnie E. John
David E. Kieras
Young-joo Jeon
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Contents
Abstract
Introduction
Comparison (1.2.3.4.)
Summary and Comparison (1.2.3.)
Conclusion
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Abstact
The GOMS model has been one of the most widely known theoretical concepts in HCI.
(Since the publication of The Psychology of HCI’)
This article compares and contrasts GOMS’s 4 family. (KLM, CMN-GOMS, NGOMSL, CPM-GOMS)
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Introduction
1.1. The Example Task
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Introduction
Goals - what the user intends to accomplish
Operators - actions that are performed to get to the goal
Methods - sequences of operators that accomplish a goal
Selection Rules - used to describe when a user would select a certain
method over the others. Selection rules are often ignored in typical GOMS analyses.
Goals vs. Operators
1.2. Definitions of 4-GOMS
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Introduction
Program Form (e.g., mark-and-delete method)
1.3. Form of a GOMS Model
Sequence Form (e.g., delete-characters method)
Advantage All procedural knowledge is visible to the analyst
Disadvantage The only way to determine… Quite time consuming
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
GOMS 중 가장 간단한 기술 .Execution time 예측 . ( 미리 정의된 행동양식의 예상시간을 비교하여 분석하는 방법 .)
e.g.)K- press a key or buttonP- point with a mouse to a target on a displayH- home hands on the keyboard or other deviceD- draw a line segment on a gridM- mentally prepare to do an action or a closely related series of primitive actionsR- the system response time (user waiting time for the system)
2.1. The Keystroke-Level Model
Architectural Basis and Constraints.Simple cognitive architecture:HIP (Human Information Processing) 의 Serial stage model 에 기반 .
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Example KLM.
2.1. The Keystroke-Level Model
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Card, Moran, and Newell GOMS.
Methods - Program form 로 표현 . (submethods, conditionals 포함 ) Operator sequence, execution time 예측
2.2. CMN-GOMS
Architectural Basis and Constraints.MHP (Model Human Processor) – parallel-stage architecture 와 HIP (Human Information Processing) 의 Simple conventional model 기반 .
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Example CMN-GOMS
2.1. The Keystroke-Level Model
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Comparison to the KLM.
2.2. CMN-GOMS
CMN-GOMS KLM
Program form (general, executable)Explicit hierarchy and goal
One-to-one mapping
Implicit hierarchyNo explicit goal
Ks and Ps mapping
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Natural GOMS Language
Program form 로 표현 . operator sequence, execution time, learning time 예측
2.3. NGOMSL
Architectural Basis and Constraints.CCT (Cognitive Complexity Theory)- simple serial-stage architecture working memory 에서 production rules 를 활성화 .
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Example NGOMSL
- Learning Time Predictions
- Execution Time Predictions
2.3. NGOMSL
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Comparison with KLM and CMN-GOMS
2.3. NGOMSL
NGOMS KLM and CMN-GOMS
Execution time
How time is assigned to cognitive and perceptual operators.
More M-like operator(Determine-position, verify)
Unobservable
operator
각 단계마다Cognition execution time 필요
(CCT 기초 )
KLM: single crude M operator
CMN-GOMS: no time (overhead)
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Cognitive-Perceptual-Motor GOMSCritical-Path-Method: (provide the prediction of total task time)
execution time 예측 - component activities 분석에 기초 - requirement of analysis level (primitive operator) : simple perceptual, cognitive, motor acts.
2.4. CPM-GOMS
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Architectural Basis and Constraints.
MHP (Model Human Processor) 에 기반 . (parallel)
Human
2.4. CPM-GOMS
Extreme expert user
Perceptual Processors Cognitive Processors
Sensory information 1st acquired recognized Physical action
Each processor : internal - serially operation, external - parallel running
deposited
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Example CPM-GOMS - Begins with CMN-GOMS model
2.4. CPM-GOMS
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
2.4. CPM-GOMS
Example CPM-GOMS
- Execution Time Predictions
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Comparison
Comparison with KLM, CMN-GOMS, and NGOMSL
2.4. CPM-GOMS
CPM-GOMS KLM, CMN-GOMS, NGOMSL
Directing mapping (CMN-GOMS)No selection rule (sequence rule)
ReasonableShorten prediction (MHP_ Expert
user)
selection rule
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Summary and Comparison
3.1. Predictions
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Summary and Comparison
3.2. Operator Time
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Summary and Comparison
3.3. Architecture Assumptions
KLM
CNM-GOMS
NGOMSL
CPM-GOMS
Easy to apply, Predicts only execution time
Predicts execution time for all subsumed task instances
Working memory and specified procedure knowledgeExplicit representation of procedural knowledgePredict Learning time
Predicts subtle execution time, overlapping patterns of activities
Simplest cognitive architecture
More complicated cognitive architecture
Powerful, unspecified multiple parallel processor architecture
Elaborated sequential architecture
Division of Information Management Engineering _User Interface Lab.
Korea University. since 1905 Conclusions
Importance of the procedures for accomplishing goals (user must learn and follow system) -quantitative predictions of procedure learning and execution time. qualitative insights into the implications of design features.
Useful tools for HCI and practical design
Expect to improved HCI models (more comprehensive, accurate, and useful)